sim.mol.data: Simulate molecular data for pathview experiment

Description

The molecular data simulator generates either gene.data or cpd.data of different ID types, molecule numbers, sample sizes, either continuous or discrete.

Usage

sim.mol.data(mol.type = c("gene", "gene.ko", "cpd")[1], id.type = NULL,
species="hsa", discrete = FALSE, nmol = 1000, nexp = 1, rand.seed=100)

Arguments

mol.type

character of length 1, specifing the molecular type, either "gene" (including transcripts, proteins), or "gene.ko" (KEGG ortholog genes, as defined in KEGG ortholog pathways), or "cpd" (including metabolites, glycans, drugs). Note that KEGG ortholog gene are considered "gene" in function pathview. Default mol.type="gene".

id.type

character of length 1, the molecular ID type. When mol.type="gene", proper ID types include "KEGG" and "ENTREZ" (Entrez Gene). Multiple other ID types are also valid When species is among 19 major species fully annotated in Bioconductor, e.g. "hsa" (human), "mmu" (mouse) etc, check:

data(gene.idtype.bods); gene.idtype.bods for other valid ID types. When mol.type="cpd", check data(cpd.simtypes); cpd.simtypes for valid ID types. Default id.type=NULL, then "Entrez" and "KEGG COMPOUND accession" will be assumed for mol.type = "gene" or "cpd".

species

character, either the kegg code, scientific name or the common name of the target species. This is only effective when mol.type = "gene". Setting species="ko" is equilvalent to mol.type="gene.ko". Default species="hsa", equivalent to either "Homo sapiens" (scientific name) or "human" (common name). Gene data id.type has multiple other choices for 19 major research species, for details do: data(gene.idtype.bods); gene.idtype.bods. When other species are specified, gene id.type is limited to "KEGG" and "ENTREZ".

discrete

logical, whether to generate discrete or continuous data. d discrete=FALSE, otherwise, mol.data will be a charactor vector of molecular IDs.

nmol

integer, the target number of different molecules. Note that the specified id.type may not have as many different IDs as nmol. In this case, all IDs of id.type are used.

nexp

integer, the sample size or the number of columns in the result simulated data.

rand.seed

numeric of length 1, the seed number to start the random sampling process. This argumemnt makes the simulation reproducible as long as its value keeps the same. Default rand.seed=100.

Value

either vector (single sample) or a matrix-like data (multiple sample), depends on the value of nexp. Vector should be numeric with molecular IDs as names or it may also be character of molecular IDs depending on the value of discrete. Matrix-like data structure has molecules as rows and samples as columns. Row names should be molecular IDs.This returned data can be used directly as gene.data or cpd.data input of pathview main function.

Details

This function is written mainly for simulation or experiment with pathview package. With the simulated molecular data, you may check whether and how pathview works for molecular data of different types, IDs, format or sample sizes etc. You may also generate both gene.data and cpd.data and check data pathway based integration with pathview.

References

Luo, W. and Brouwer, C., Pathview: an R/Bioconductor package for pathway based data integration and visualization. Bioinformatics, 2013, 29(14): 1830-1831, doi: 10.1093/bioinformatics/btt285

Examples

Run this code

#continuous compound data
cpd.data.c=sim.mol.data(mol.type="cpd", nmol=3000)
#discrete compound data
cpd.data.d=sim.mol.data(mol.type="cpd", nmol=3000, discrete=TRUE)
head(cpd.data.c)
head(cpd.data.d)
#continuous compound data named with "CAS Registry Number"
cpd.cas <- sim.mol.data(mol.type = "cpd", id.type = "CAS Registry Number", nmol = 10000)

#gene data with two samples
gene.data.2=sim.mol.data(mol.type="gene", nmol=1000, nexp=2)
head(gene.data.2)

#KEGG ortholog gene data
ko.data=sim.mol.data(mol.type="gene.ko", nmol=5000)

Run the code above in your browser using DataLab

Data engineering and BI courses are free!